Martin Splitt from Google clarified that if Google sees in the initial HTML a noindex; it won’t be able to process any meta robots changes or information from within your JavaScript. Martin said “I tested if we process changes to the meta robots in JS when the initial HTML has a noindex. We don’t.”

Here is what he posted on Twitter:

This does not mean Google won’t read and process the other HTML, it may. He said “As you can see, we got the HTML from the crawler anyway, so we *may* process it anyway. Or not. I don’t know and I’m sure the answer could be “it depends””

Will this be the way Google handles this forever? Probably not. But who knows.

Of course, you can test this yourself over and over again if you’d like.

Forum discussion at Twitter.